Troubleshooting Alternator Wiring

[ecurbruce]

Finally some decent test results;(from neg. side battery)
Amps draw- ignition on engine not running=0.32 amps
-------------- ignition on not running h-lamp on=5.6 amps**

Engine idleing h-lamp off=0.80-1.30 amps*
Engine idleing h-lamp on=6.1 to 6.6 amps**

These readings are with charging system disconnected, and 10 amp charger on battery on at all times.

*this reading levels out closer to one amp at higher RPM's.

**This includes halogen 60 watt h-lamp, LED tail lamp.

Ran it all twice some three times to verify...

Bruce

[DewCatTea-Bob]

[quote= ecurbruce ...
" I'll confirm with you before winding, so we're on the same page... "

____ Alright,, as it's pretty-hard to be on the same-page, when the page YOU were on doesn't even exist anywhere here.



" We now have a pair of #22's, let's go ahead and test them, "

____ I was before left under the impression that you had already sent them to Bill.
Hopefully Bill will keep us regularly updated on the various test-setups that he wishes to try-out.



" we're never too far along to go back and rewind something else we need. "

____ I'm still not sure if YOUR chosen winding-option was purposely-planed or the results of confusion between past considered winding-options.
I recently before had assumed that you had purposely-chosen to jump-ahead & go-with plan-B (in the utmost way).
Are you now saying that you'd still be willing to also do other winding-options as well, (and possibly also still consider more-suitable/rather-balanced power-production options) ?



" I can think of a couple of things;
My meter only shows one decimal place, "

____ Yes, I was aware of that likely shortcoming, which is why I 'averaged' the two calculation-methods together (so that the resulted-figures wouldn't become too far outside of the ballpark).


" so the .2 obms could be as high as .29, or as low as .2 ,,, or low as .15, and high as .25... I don't know? "

____ Yes that's too-bad that ohm-meters don't go at-least one-more decimal-digit.
According-to the-chart though, (and along-with your updated-data),, your ".2" meter-reading actually ought-to really be '.297' ohms ! _ And-also, your ".3" reading (for one-length within the large-spool) ought-to be '.412' ohms.
So unless your chosen wire is pure-silver, I would rather suspect that YOUR ohm-meter is fairly inaccurate (more-so than the chart-table).
__ Is the ohm-meter that you got your .2 & .3 readings from, the very-same meter that you used to get the 1.4 ohm-reading which you previously posted for the small-spool winding of #24-gauge ?



" The outermost winding is 3-3/8 inch,
The bobbin measures 2-3/4 inch, circumference, "

____ With THOSE more-detailed (and slightly larger) dimensions, I figure the 344 loop-turns must be very-near to 87-feet,9-inches (in TOTAL-length for both power-coils) ! _ And that length is pretty-close to the grand-total winding-length of all-SIX stock power-coils that Bill had tested while arranged in three-pairs connected in parallel (to get at-least 80-watts, [as I recall]).
__ So NOW, we can be fairly quite SURE that your rewound-pair will pretty-much match the same results that Bill had attained with his three-sections in parallel setup !
Which means that such a plan-B stator should produce well-over 240-watts !
So you'll probably be able to power your Z.bulb's high & low beams both at once (and still keep your battery well charged) !



" And you've always been a little suspect of that chart, "

____ I've actually been pretty-much trusting it,, as the only aspect that I'm sure is quite misleading about it, is it's listed current-capacities stated for each of the various wire-gauges.



" you may be right about it being a little off??? "

____ Only in regards to 'current-capacity', in-which it's certainly MORE than just a "little off" !
Cuz most-certainly gauge-#22 can fairly easily cope with well-over just ".92" amps ! _ I'd expect that size of wire to-be able to pass at-least 5-amps without any issues.
__ How did you happen-to find that chart-table ?



Enlightened-Cheers,
DCT-Bob

[ecurbruce]

Bob says;"
Are you now saying that you'd still be willing to also do other winding-options as well, (and possibly also still consider more-suitable/rather-balanced power-production options) ?

I'm still open to anything, nothing is set in stone in my opinion.
Let's see what happens after some testing and rewinding...

Bob says;"____ I was before left under the impression that you had already sent them to Bill. "

Oh, yea, was right out the door with them, when it seems we needed more dialogue.?
He'll have them real soon.

Bob says"
__ Is the ohm-meter that you got your .2 & .3 readings from, the very-same meter that you used to get the 1.4 ohm-reading which you previously posted for the small-spool winding of #24-gauge ? "

Yes, same one.


Bob says"__ How did you happen-to find that chart-table ? "

Typical "Google" search, was actually only looking for wire sizes, the other elements in that chart were just a bonus...
About the current limit specs, aren't they relative to the voltage, and maybe this is not at 12 volts?


Bob, did you see the test results of my electrical system from thursday afternoon?

Bruce

[DewCatTea-Bob]

[quote= ecurbruce ...
" I'm still open to anything, nothing is set in stone in my opinion. "

____ Well you ought've realized the reasoning for my inclination to assume that you had chosen to jump-ahead to plan-B, (considering your posts placed since your 10-day leave).
__ I've been thinking that besides creating an ultimate-stator with 3-pairs of 'double-wound' -(not 'dual-wound' like Ducati's), power-coils,, we could also create a rather balanced-stator with just the 2-pairs of power-coils. _ (Which means I'd then send you both of my n-c.6-ploe stators for rewinding with OTHER winding-options.)



" Let's see what happens after some testing and rewinding... "

____ But of-course, as that's the road that must logically be taken before commitment to any particular finalized-option.



" Oh, yea, was right out the door with them, when it seems we needed more dialogue.? "

____ Indeed, ditto over-here at my-end as well,, due-to:

Bill, I'm sending them your way, you have a PM.

And considering the work of rewinding, it wasn't unsensible to have gone-ahead & sent them out on their way.

__ Is the ohm-meter that you got your .2 & .3 readings from, the very-same meter that you used to get the 1.4 ohm-reading which you previously posted for the small-spool winding of #24-gauge ? "

"Yes, same one. "

____ I believe that chart-table and your-meter turned-out pretty-close for the actual-length of gauge-#24 which you later measured -(" the small spool I unwound held 56 feet, 8 inches of wire! "). _ As the calculated resistance for that length is 1.454-ohms, and your meter-reading was "1.4".



" Typical "Google" search, was actually only looking for wire sizes, the other elements in that chart were just a bonus... "

____ I'll try a different search-engine and see if I can find another chart-table that lists real-world current-capacities for at-least the wire-gauge sizes which we may be interested in.
UPDATE - Here's a link to a page with a table-list which indicates that gauge-#22 can handle 7-amps... http://www.powerstream.com/Wire_Size.htm



" About the current limit specs, aren't they relative to the voltage, and maybe this is not at 12 volts? "

____ In such a (actually unrealistic) consideration, I'd expect such charts would then have-to state the current @ whatever-voltage, in-order to make any sensible-sense.
ACTUALLY however (realistically!), an established-amount of 'current' through a conductor is the same current-amount -(amperage) regardless of whether being pushed-along by just .12-volts or even 12k-volts ! _ As 'current' always remains being the very-same as that which it always is, regardless of any associated voltage !
In other-words, 1-amp of current is always the very-same amount of electron-flow, no mater the amount of voltage that established it.



" Bob, did you see the test results of my electrical system from thursday afternoon? "

____ Yes, but I no-longer felt much need for that data, since I had been led to believe that you had become unconcerned with any attempt to attain a rather balanced system, and-thusly little need for matching plan-A's prime-winding to be more balanced accordingly with that data.
__ I'll soon work-on that stuff though, anyhow.


____ I'm going to add more to this post, later.
-B

[wcorey]

Received the coils today.
Unfortunately very disappointing test results.

The large rewound one weighs 52 grams, the stock one 32.
The small rewound one weighs 36 grams, the stock one 20.
Fat little things, both just barely fit on the stator between the core/post and the mounting plate.
Must have been some careful planning there.

Was a bit nostalgic dusting things off and setting it all back up for the test runs, lol.

Was evident pretty quickly that this is a dead end or something is off, so just ran them through a few quick paces to get an idea where we're at.
I didn't spend a lot of time on it or bother to write anything down and figures are just approximate averages.

Certainly some odd combinations of numbers but I tried a few redundant tests as well as different combinations/configurations (loads, rpm's, connections, etc)
for each basic test and got fairly consistent results.



3400rpm, 11ohm load (because that achieved around 12vdc), generic full wave bridge rectifier.
Small one, 9w
Large one, 13w

Both coils running simultaneously on their own rec's with combined rec outputs only got 15w...?
Barely got a noticeable glow from a headlight, did fire the brakelight up well enough though.

3400rpm, 11ohm load, mosfet r/r.
Small one, 16w
Large one, 18w

Combining both coils to the r/r inputs got 18w, no big surprise there with the considerable mismatch, I guess.


I think next obvious thing to try is taking the two windings out of parallel and running them both separately and then together as one continuous series.
And if that works out then both of those coil/setups in series, then parallel...


Ok, I tried connecting the two windings (on one coil) in series, picked up 2-3 more watts across the board than the previous figures.
Both coils in series/series (with each other), most I got was 26w.

I guess that's about all I've got here. Any suggestions/requests before they go back to Bruce?

[DewCatTea-Bob]

[quote= wcorey
" Received the coils today. "

____ Bad-timing, as I've been off-line for over the past day !



" Unfortunately very disappointing test results. "

____ I hadn't expected that you would get-started so quickly before announcing that you've received the pair of power-coils and first asked for suggested test-setups.
Still though, I had expected that you'd still recall what-all needs to be done in order to find & discover whatever the MAX.power-output actually is.



" The large rewound one weighs 52 grams, the stock one 32.
The small rewound one weighs 36 grams, the stock one 20. "

____ Kind-of strange that these figures don't seem to sensibly inter-correlate well.



" Was evident pretty quickly that this is a dead end or something is off, "

____ I'm sure it's rather that "something is off " !



" Certainly some odd combinations of numbers but I tried a few redundant tests as well as different combinations/configurations (loads, rpm's, connections, etc)
for each basic test and got fairly consistent results. "

____ I'm sorry that it seems you had wasted your time, but it was no-doubt good for you to get some fresh/recent-practice under your belt, before continuing-onward.



" 11ohm load (because that achieved around 12vdc),
Small one, 9w
Large one, 13w

____ WHEN did that (overly SMALL !) load ever result with 12-volts ?
__ That (rather irregular) amount of resistance is akin-to as-if taking a (wrong-way) step towards the equivalent of connecting a single-element LED.pilot-light to the combined-output of all the power-generators of the Hoover-Dam and with such a puny-load circuit-setup still somehow expect to be enabled to measure all the capable power-output that's possibly available from that whole/entire awesome power-plant ! _ So considerably more obvious,, measuring the power consumed by such a miniscule load, certainly isn't going to reveal how much power the power-plant can possibly actually provide !
__ It must be understood that in order to measure the MAXIMUM-possible power-output, you thus-then need to connect a suitable-LOAD that's really actually capable of consuming all the power that's possibly available to be consumable !! _ So ya got-to realize that a load as light as 11-ohms can't accept all the power-juice that could possibly be available from a power-winding that has it's-own impedance-level set way-below 11-ohms.
__ At this point, we don't need to be so concerned about exactly what the specific voltage & current values seem to be,, but rather, merely what wattage-level those two combined power-aspects equate to.



" Both coils running simultaneously on their own rec's with combined rec outputs only got 15w...?
Barely got a noticeable glow from a headlight, did fire the brakelight up well enough though. "

____ Don't know for-sure what you're actually doing to have gotten such results, as there's a number of things that could've been responsible.
__ For now, I'm going to ASSUME that your (rather lacking) wording is meant to indicate that you (separately) took each power-coil's double-windings and left them side-by-side & end-to-end (as-if same as a single-winding) and-then connected to a dedicated full-wave rect.block,, and then-next parallel-combined the outputs of both rect.blocks to a load (which evidently was-not 'matched'), to obtain a "15w" result ?
__ Did you try measuring the DC.output of EACH power-coil/rect.block circuit, separately ?
____ Also, did you inspect the two power-coils to check & see that Bruce had remembered to correctly wind them both in opposite-directions from one-other ?



" 11ohm load, mosfet r/r. "
Small one, 16w
Large one, 18w "

____ That test is one I was going to ask for LATER, to see if your magic-R-R.unit would also do it's-thing on a rather straight-type winding as well.
But that's for AFTER matters-at-hand have first been sorted-out.



" Combining both coils to the r/r inputs got 18w, no big surprise there with the considerable mismatch, I guess. "

____ Well of-course directly-connecting the large power-coil to the small power-coil in parallel, would of-course thus-then certainly allow the small-one to rob power from the large-one BEFORE whatever leftover-power is then-next rectified & measured !



" I think next obvious thing to try is taking the two windings out of parallel and running them both separately and then together as one continuous series. "

____ Indeed those test-ideas are just two tests which I would've asked if you would care to try-out, only AFTER the main/intended-test had been finished.
The main/intended test-work was to check for the expected increase of power (due-to increased winding-length), compared to one of the paired-sections that you had before done (with stock coil-windings).
That means that the double-windings on each of the two power-coils are to be kept/left in parallel (just as-if a SINGLE-winding), and-then the two power-coils connected to each-other in SERIES !
Then with both power-coils fitted onto neighboring stator-core fingers,, the large-coil's upper-left output-leads should be connected to an AC.input of a rect.block, while the small-coil's (opposed) lower-right output-leads are connected to the rect.block's opposite AC.input. _ And then the DC.output of the rect.block should be connected to a 'MATCHED' load !
__ To determine a "suitable" load that's 'matched' to the combined power-winding, we first have-to start-out by measuring the resulted resistance of the combined power-winding !
(Measuring is best but, if we go by what Bruce has already measured for one winding of the small power-coil, [which was merely-just .2-ohms], then those two identical power-windings connected-together in series should add-up to .4-ohms,, which thusly means that the connected test-load needs to be as low as .4-ohms, in order to PROPERLY-determine the ACTUAL max.power [of THAT combined-winding]!)
__ So measure that combined-winding with an ohm-meter,, and then we start-out with a load that has the SAME resistance-value.
Now it won't be quite as simple as merely-just that,, because as RPM.frequency climbs, impedance then becomes a greater & greater factor. _ So further-increased resistance-increments will need to be added in steps until the measured max.power has become peaked (and still further increases [in load-resistance] then-next begin leading-to decreases in measured-power). _ So when the 'peak' of power has finally been pinned-down & precisely determined, then the amount of load-resistance found (to discover whatever the MAXIMUM power actually is), is the amount of 'impedance' of the power-winding (at whatever the test-RPM).
Only THEN,, when the test-load has become 'MATCHED' to the power-winding, can the MAX.power of the power-winding be determined !!



" And if that works out then both of those coil/setups in series, then parallel... "

____ Right, then that too, next. _ But didn't you state above that you've already tried the power-coils in parallel arrangement ?



" I tried connecting the two windings (on one coil) in series, picked up 2-3 more watts across the board than the previous figures. "

____ You failed to mention if the test-load was still as much as 11-ohms,,
so if-so, then the reason for the measured power-increase, is due-to having somewhat more-closely matched the resulted power-winding to the connected test-load.



" Both coils in series/series (with each other), most I got was 26w. "

____ At-least a couple things to get straight here...
__ First,, I gather this to mean that you've put both windings of both power-coils -(ALL-4 separate power-windings) chained-together all in SERIES-fashion !?
Assuming so, then was your test-load's resistance not perfectly 'matched' to the resulted resistance of the combined power-winding ? _ Cuz, as not doing so, (as my LED.vs.Hoover-Dam analogy has meant to make obvious), will very quite likely lead-to a rather MEANINGLESS power-assessment !
__ Next,, are you quite sure that you had all four power-windings properly connected so that ALL their polarities were arranged in the very-same direction (so that all their power would properly ADD-up together) ?
Cuz if Bruce hadn't thought to wind the two power-coils oppositely, then YOU'd have-to realize that & then be extra-careful to make-sure that you connect the two power-coils in PROPER-series so that they're not possibly tending to cancel one-another.
(And also, connecting them wrong PURPOSELY, may also be a test worthy of trying-out, [just to be doubly sure].)



" Any suggestions/requests before they go back to Bruce? "

____ YES, as there's lots-more to get all ironed-out before we're done, (sorry to say) !
__ Bottom-line here/now at this point, is that "11 ohms" is WAY-TOO F-n "off" to tell us anything of much use ! _ I don't understand how you could possibly think that that-amount of load-resistance could determine anything at-all, after you had not long ago seemed to properly realize that Bruce's #24-winding's resistance of 1.4-ohms was TOO-much !?!
Cuz certainly if you had actually tested rather with a '1.1-ohm' test-load, THEN no-doubt you would've also gotten MUCH-higher power-results !
But the test-load for the main/intended test, should be even lower ! ...
Bruce measured the resistance of the windings of the small-coil to be .2-ohms each, and that of the large-coil to be .3-ohms each,, so that means that the test-load ought-to start-out at about .2-ohms (and-then work-upwards from there, until the 'peak' is found).
So test-load increments of .05-ohms would be in order,, and to get such small incremental-steps, you'd probably need to find a length of 12 to 16-gage wire that measures .2-ohms and cut 75% of it
away (so that it can be added as needed to dial-in the test-load).


Hopeful-Cheers,
-Bob

[ecurbruce]

Bill says;"I guess that's about all I've got here. Any suggestions/requests before they go back to Bruce?"

Bill, when all of Bob's testing has been exhausted, would you concider unwinding two layers of both those spools, then run the most conclusive test on them again to see the result of slightly decreasing the winding size? Those windings are the MAXIMUM size, not necesarily the "best size", that may give an indication.
Thanks,
Bruce

[wcorey]

First off, I realize my reporting of results is in rather shoddy form. My last post may be a little confusing because I edited in the last part later on as I went along the next day, (for some reason the site no longer indicates that posts have been edited).
I haven't had much time for detail work and after a few quick tests proved to be so underwhelming, I just didn't bother to do any meticulous/methodical testing at all but rather banged out a bunch of combinations hoping to find the magic bullet (and not wanting to waste time on potentially 'useless' stuff until I heard back from the two of you).

Not even sure where to begin as I probably did 50 or more separate test setups in quick succession to get a general feel for what's what.
Methodology was consistent with past testing so I assumed that some of the blanks here would have been filled in by our mutual past experience.
I did a quite a large amount of quick 'n dirty setups, many mistakes are possible but results were pretty predictable so I doubt that mattered much.
Initially tested down to one ohm and up to 15 or so and found the wattage outputs to be fairly close.
As I've always done in the past, put a 55w headlight on some of the setups as a 'control' to see what happened there (not much).
Bruce always wants to know around where 12v happens so that's why I reported the results at 11 ohms, as opposed to 'peak output'.

With all those different loads and all the different series, parallel, single winding, double winding, single coil, double coil, etc arrangements, NOTHING stood out with any singularly impressive gain over the majority of others. Hence my lack of bothering to report in with any great detail until I got some input from you guy's.
If/when I go down to sub-ohm loads, I'm guessing we'll see something like less than 2v, 10-12amp at 3400rpm on the large coil alone.

The coils didn't need to be wound in any particular direction as switching the 'polarity' is effectively the same as switching the direction.
If I wire anything 'backwards'/'out-of-phase' it's pretty obvious as I generally get close to no output.

I'm off to real work so must continue this later, sorry to leave y'all hanging...

[DewCatTea-Bob]

[quote= ecurbruce ...
" would you concider unwinding two layers of both those spools, then run the most conclusive test on them again to see the result of slightly decreasing the winding size?
that may give an indication. "

____ We might learn something from doing that experiment but,, removing any of the winding-length will reduce the amount of flux-line cutting,, and also lower the winding-resistance still further yet, which may-not be a good-thing since optimum power-transfer only occurs when the power-winding is impedance-matched with the load.


Hopeful-Cheers,
-Bob

[DewCatTea-Bob]

[quote= wcorey ...
" My last post may be a little confusing because I edited in the last part later on as I went along the next day, "

____ Not really so confusing, since your fairly specific wording fortunately somewhat clues the reader that some further-progression has transpired.



" (for some reason the site no longer indicates that posts have been edited). "

____ For over a couple of years, I've noticed with a fair number of posts by various members that many edits have occurred without being automatically noted, so it seems to be a hit-or-miss type of occurrence that can't be depended upon to always be noted.



" Not even sure where to begin as I probably did 50 or more separate test setups in quick succession to get a general feel for what's what. "

____ Well at-least it's pretty-nice to know that you're so enthusiastic about learning so-much !
__ I assume that you always tested the pair of coils with them loaded onto stator-core fingers that were next-door neighbors ? ...
I think it would be a worthy test to try the two power-coils loaded onto the core-fingers located at the 9 & 3 o'clock positions, with a series-connection made between those two power-coils.



" Methodology was consistent with past testing so I assumed that some of the blanks here would have been filled in by our mutual past experience. "

____ Well yeah but, not so considerate for all the OTHER readers.
I'll try to help that-cause (and help spare you from as much extent), by posting more detailed wording of what (I think) you mean to state. - (So correct me if I ever screw-up on that endeavor.)



" Initially tested down to one ohm and up to 15 or so and found the wattage outputs to be fairly close. "

____ As you load-tested with resistance-values that continually moved further away from .2-ohms, the resulted wattage should've likewise continually diminished.



" If/when I go down to sub-ohm loads, I'm guessing we'll see something like less than 2v, 10-12amp at 3400rpm on the large coil alone. "

____ Certainly would like to see if your prediction is very close !
You could hook-up five 1-ohm resistors in parallel, to achieve the .2-ohm test-load,, or perhaps a spool of wire may provide such a resistance-level.
It shouldn't be too awfully difficult for you to come-up with such low-resistance test-loads.



" The coils didn't need to be wound in any particular direction as switching the 'polarity' is effectively the same as switching the direction. "

____ Yeah, ya would think-so, wouldn't ya !
But didn't we discover that Ducati had actually bothered to wind the small power-coils oppositely from the direction chosen for the large ones ?



" If I wire anything 'backwards'/'out-of-phase' it's pretty obvious as I generally get close to no output. "

____ Even-so however, there should still be some 'leftover' power to be measured, (as the large-coil should overpower the smaller one).



" must continue this later, sorry to leave y'all hanging... "

____ NO-problem at all, you're just fine Bill ! _ Please don't feel any pressure at all, as you're certainly entitled to take any break for ANY reason you like !!




____ To once-again cover exactly-what test is of most concern for Bruce's pair of power-coils (with double-lengths of #22-wire power-windings),,
since that paired-set -(one small & one large) is best suited for plan-B, that pair should be tested connected-together in SERIES and arranged in-place on the stator-core so that when one of the power-coils is aligned with a North-magnet, the other-coil is then aligned with a South-magnet. - (Cuz if both coils get aligned with the same S/S or N/N polarity, then the magnetic-fields will-not be enabled to properly combine for FULL-effect, [and switching-around the polarity of the electrical-connections in attempt to compensate, will-NOT help] !)
__ For this main/intended-test,
the fact that both of the two power-coils each have two separate power-windings, is to be IGNORED,, as those double-windings are to be left in parallel and dealt-with as-if mere single-windings !
However, since the power-windings (of the two power-coils) are-not exactly the very-same length,, the pair of power-coils are intended to be connected-together ONLY in series-fashion ! - (With that series-connection between the two power-coils being properly polarized, of-course.)
Then the remaining two/(2-pairs) output-leads -(the 0-phase of the large-coil, and the 180-phase of the small-coil) are to be connected to a std.full-wave bridge-rectifier -(rect.block), and
it's output connected to the test-load.
__ The test-load's resistance should be no lower than that of the combined-winding -(of-which IT's resistance should be measured). _ Going-by Bruce's given measurements, the resistance-level for the 'combined' power-winding (that's actually arranged as parallel/series/parallel), should be about .25-ohms total,, so that's about the level where the test-load should start-out at. _ And from there, (as RPM/impedance increases), the load-resistance should be incrementally increased until the peak-power has been reached (and further increments of load-resistance then only yields further diminishing amounts of power).
(The found-difference between the discovered optimum load-resistance and the .25-ohms [of the combined-winding], is the amount of 'impedance' caused by the frequency of the AC at the test-RPM. - [So the actual-optimum load-resistance will be equal to the combined-winding's .25-ohm resistance-level PLUS it's X-ohm impedance-level.] )
Obtaining a test-load resistance-level of just .25-ohms, could be achieved by arranging four 1-ohm resistors all in parallel,, (while just three would achieve .33, and just two would be .5-ohm).
__ Finally, in order to determine the max.wattage that the tested power-winding is capable-of,, the measured 'amperage' flowing-through the test-load multiplied with the 'voltage' dropped across the test-load, will result with the 'wattage' that's available for consumption !


Hopeful-Cheers,
D.Bob